Struct AgentContext 

pub struct AgentContext<T> {
    pub llm: LlmClient,
    /* private fields */
}

Context provided to agent handlers.

This gives agents access to LLM inference and the ability to emit results.

Fields

llm: LlmClient

LLM client for inference calls.

Implementations

impl<T> AgentContext<T>

pub fn emit(&mut self, value: T) -> SageResult<T>

where T: Clone,

Emit a value to the awaiter.

This should be called once at the end of the agent’s execution. Calling emit multiple times is a no-op after the first call.

pub async fn infer<R>(&self, prompt: &str) -> SageResult<R>

where R: DeserializeOwned,

Call the LLM with a prompt and parse the response.

pub async fn infer_string(&self, prompt: &str) -> SageResult<String>

Call the LLM with a prompt and return the raw string response.

pub async fn receive<M>(&mut self) -> SageResult<M>

where M: DeserializeOwned,

Receive a message from the agent’s mailbox.

This blocks until a message is available. The message is deserialized into the specified type.

pub async fn receive_timeout<M>( &mut self, timeout: Duration, ) -> SageResult<Option<M>>

where M: DeserializeOwned,

Receive a message with a timeout.

Returns None if the timeout expires before a message arrives.